High Resolution Analysis and Phylogenetic Network Construction Using Complete mtDNA Sequences in Sardinian Genetic Isolates Cristina Fraumene,* Elise M. S. Belle, Loredana Castrı`,à Simona Sanna,§ Gianmaria Mancosu,* Massimiliano Cosso,* Francesca Marras,* Guido Barbujani, Mario Pirastu,*§ and Andrea Angius*§ *Shardna Life Sciences, Pula (Cagliari), Italy; Dipartimento di Biologia, Universita` di Ferrara, Ferrara, Italy; àDipartimento di Biologia Evoluzionistica e Sperimentale–Sezione di Antropologia, Universita` di Bologna, Bologna, Italy; and §Istituto di Genetica delle Popolazioni, CNR, Alghero (Sassari), Italy For mitochondrial phylogenetic analysis, the best result comes from complete sequences. We therefore decided to sequence the entire mitochondrial DNA (mtDNA) (coding and D-loop regions) of 63 individuals selected in 3 small Ogliastra vil- lages, an isolated area of eastern Sardinia: Talana, Urzulei, and Perdasdefogu. We studied at least one individual for each of the most frequent maternal genealogical lineages belonging to haplogroups H, V, J, K, T, U, and X. We found in our 63 samples, 172 and 69 sequence changes in the coding and in the D-loop region, respectively. Thirteen out of 172 sequence changes in the coding region are novel. It is our hypothesis that some of them are characteristic of the Ogliastra region and/ or Sardinia. We reconstructed the phylogenetic network of the 63 complete mtDNA sequences for the 3 villages. We also drew a network including a large number of European sequences and calculated various indices of genetic diversity in Ogliastra. It appears that these small populations remained extremely isolated and genetically differentiated compared with other European populations. We also identified in our samples a never previously described subhaplogroup, U5b3, which seems peculiar to the Ogliastra region. Introduction Achilli et al. 2004; Howell et al. 2004; Palanichamy Mitochondrial DNA (mtDNA) analysis is considered et al. 2004; Rajkumar et al. 2005), represent the best an essential tool for studying human population structure, possible solution for phylogenetic analysis (Richards and origins, migration patterns, and demographic history, given Macaulay 2001; Kivisild et al. 2006). However, data for its polymorphism, its matrilineal mode of descent, and its mtDNA polymorphisms in different human populations lack of recombination (Torroni et al. 1996). The first com- are still limited and only few haplogroups are based on plete sequence of human mtDNA, the Cambridge Reference mtDNAs complete sequence (Finnila and Majamaa 2001). Sequence (CRS), was published in 1981 (Anderson et al. Most important, several reports associate some diseases 1981) and was recently revised (rCRS, Andrews et al. 1999). with specific mtDNA haplogroups (Brown et al. 1997; The fact that mutations accumulate sequentially along Kalman et al. 1999; Chinnery et al. 2000; Ruiz-Pesini maternal lineages allows associating many of them with dif- et al. 2000; Moilanen et al. 2003; Herrnstadt and Howell ferent geographical regions of the world (Ingman et al. 2004; Mancuso et al. 2004; Pyle et al. 2005), and it is there- 2000; Herrnstadt et al. 2002). mtDNA sequence varia- fore fundamental to increase our understanding of mtDNA tions can thus be used to construct phylogenetic networks haplogroup (Rose et al. 2001; Niemi et al. 2003, 2005). (Bandelt et al. 1999), displaying the relationships among Interpopulation comparisons and phylogenetic tree sequences and estimating the time of appearance of muta- construction through mtDNA studies can be useful for tions associated with each haplotype. the characterization of populations with unusual genetic fea- Historically, the first mtDNA polymorphisms used in tures (Tolk et al. 2000; Finnila and Majamaa 2001; Larruga human phylogenetic studies were identified in the noncoding et al. 2001; Meinila et al. 2001). mtDNA analysis of differ- or D-loop region, containing the hypervariable segments ent Sardinian populations revealed specific genetic charac- (HVS). Accurate phylogenetic networks for European teristic and a variable degree of subpopulation homogeneity mtDNAs were constructed using HVS-1 and HVS-2 se- within the island (Workman et al. 1975; Piazza et al. 1988; quence data (Richards et al. 1998; Helgason et al. 2000), Morelli et al. 2000). However, so far no studies were carried complemented with coding-region restriction fragment out in eastern Sardinia using complete mtDNA sequences. length polymorphisms (RFLP) used to define mtDNA hap- In the present article, we present the analysis of 63 logroups (Torroni et al. 1996; Macaulay et al. 1999). High complete mtDNA sequences of samples coming from 3 dis- mutation rates, variability in site substitution rates, homopla- tinct villages of a Sardinian region, Ogliastra, characterized sic sites, parallel mutation events, and reversion make the D- by both genetic and environmental homogeneity. This area loop evolution complex and the phylogenetic analysis sub- encompasses 23 small isolated villages, whose founders ject to error (Bandelt et al. 2002; Dennis 2003; Forster 2003). presumably derived from the same original gene pool, with Complete mtDNA sequences, only recently becoming avail- scant genetic exchanges with the rest of Sardinian areas able (Ingman et al. 2000; Finnila et al. 2001; Maca-Meyer recorded during the last 400 years of parish and historical et al. 2001; Torroni et al. 2001; Herrnstadt et al. 2002; records. Indeed, it is thought that geographic and cultural barriers have exerted a strong isolating effect in this part Key words: complete mtDNA sequences, genetic isolates, phyloge- of Sardinia. Geographical isolation, small population size, netic network, subhaplogroup, genetic drift, founder effect. high endogamy, and inbreeding are expected to lead to in- E-mail: [email protected]. creased genetic differentiation among subpopulations as Mol. Biol. Evol. 23(11):2101–2111. 2006 doi:10.1093/molbev/msl084 a consequence of founder effect and genetic drift (Angius Advance Access publication August 10, 2006 et al. 2001; Fraumene et al. 2003). Ó The Author 2006. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected] 2102 Fraumene et al. Table 1 Maternal Genealogical Lineages of the 3 Villages. ‘‘Year’’ Indicates the Date of Introduction of Maternal Lineages in the Village.‘‘Progeny’’ Indicates the Total Number of Living Descendants. ‘‘Hapl.’’ Designates the Mitochondrial Haplogroup Talana Urzulei Perdasdefogu Founder Year Progeny Sample Hapl. Founder Year Progeny Sample Hapl. Founder Year Progeny Sample Hapl. 1 1600 97 2545T H 1 1780 91 00136U H 1 1750 76 0002P H 2 1600 104 2175T H1 2 1570 35 00207U H1 2 1680 73 0218P H1 3 1623 108 2554T H1 3 1805 40 00097U H1 3 1680 64 0305P H1 4 1684 95 2798T H1 4 1572 409 00004U H3 4 1685 36 0695P H1 5 1720 33 2472T H1 5 1585 78 00070U H3 5 1690 198 0399P H1 6 1800 221 2579T H1 6 1790 19 00307U H3 6 1879 45 0157P H1 7 1572 104 2702T H3 7 1865 21 00120U H3 7 1640 60 0437P H3 8 1585 94 2060T H3 8 1890 23 00271U H3 8 1665 27 0019P H3 9 1600 34 2477T H3 9 1763 140 00062U H4 9 1695 182 0009P H3 10 1640 13 2465T H3 10 1600 22 00257U J1c 10 1700 35 0054P H3 11 1760 170 2538T H3 11 1610 217 00134U J2b 11 1851 23 0066P H3 12 1770 45 2662T H3 12 1590 64 00100U T2b 12 1650 101 0006P PreV 13 1790 33 2008T H3 13 1770 46 00260U T2b 13 1870 34 0540P J2a 14 1895 16 2282T J2b 14 1776 63 00105U T2b 14 1680 212 0349P K1a 15 1740 109 2038T T2 15 1660 66 00050U U5b1 15 1884 32 0208P K1a 16 1662 128 2096T U5b3 16 1780 33 00191U U5b2 16 1610 95 0259P T2 17 1550 130 2010T V 17 1785 183 00102U U5b2 17 1854 38 0185P T2 18 1590 45 00368U U5b3 18 1660 34 0288P U1a 19 1614 106 00032U U5b3 19 1725 141 0008P U1a 20 1615 63 00187U U5b3 20 1650 47 0044P U5b3 21 1650 104 00213U U5b3 21 1831 34 0176P U5b3 22 1904 24 00343U U6 22 1800 48 1074P X2b 23 1710 218 214P-358P X2 Genealogical records systematically kept since the for Urzulei, 22 genealogical lineages account for about 17th century allow the careful and accurate reconstruction 90% present-day living inhabitants (2,091 residents and of genealogies for each village. We created a relational da- nonresidents), and for Perdasdefogu, 23 genealogical line- tabase and developed specific tools to access these data eas- ages account for about 79% present-day living inhabitants ily and reconstruct complete genealogical trees for up to 16 (2,340 residents and nonresidents). During the last 50 years, generations in a rapid manner (Mancosu et al. 2003, 2005). Perdasdefogu has undergone substantial immigration be- It is, therefore, possible to also reconstruct accurately all the cause of a military base nearby, but, by our sampling maternal lineages present in the 3 villages all the way back method, the maternal lineages immigrated in the village to the 17th century. We analyzed one or more individuals after 1950 were not considered. from each founder maternal genealogical lineage. Informed consent was obtained from each individual Our study allowed us to monitor segregation and se- and all samples were collected in accordance with the lection in the evolution of these lineages. Complete mtDNA Declaration of Helsinki. sequences permitted us to evaluate genetic differentiation of each village in comparison to other European sequences. DNA and Sequence Analysis Genomic DNA was isolated from 5 ml of peripheral Materials and Methods blood as previously described (Ciulla et al. 1988). Whole mtDNA was amplified using 24 partially overlapping frag- Samples Selection ments, each ’800 bp in length (Rieder et al. 1998). Poly- We sequenced complete mtDNAs of 63 individuals merase chain reaction (PCR) primers were designed to from 3 small villages within Ogliastra region (Eastern provide ’200 bp overlap between neighboring fragments.